+ Reply to Thread
Results 1 to 3 of 3

Thread: random sampling versus choice based sampling

  1. #1

    random sampling versus choice based sampling




    Hi everyone,
    I have got a question about sampling. I have collected all the available firms from a database and the resultant dataset is an unbalanced sample of 10% failed firms and 90% active firms. My understanding is that this is close to random sampling.

    In my research area, some earlier studies used balanced or matched sample of 50% failed and 50% active. Is this the choice based sampling?

    Regards
    Tim

  2. #2
    Points: 3,631, Level: 37
    Level completed: 88%, Points required for next Level: 19
    staassis's Avatar
    Location
    New York
    Posts
    226
    Thanks
    2
    Thanked 41 Times in 39 Posts

    Re: random sampling versus choice based sampling

    Any balancing is fine for the validity of the analysis, as long as you sample the firms objectively (not just take those which come to mind). However, a split close to 50%-50% is likely to delver the maximum statistical power for any given sample size.

  3. The Following User Says Thank You to staassis For This Useful Post:

    t.wang21@yahoo.com.au (06-30-2017)

  4. #3
    Points: 7,325, Level: 56
    Level completed: 88%, Points required for next Level: 25

    Posts
    116
    Thanks
    35
    Thanked 0 Times in 0 Posts

    Re: random sampling versus choice based sampling


    Thanks Staassis. While 50%-50% may deliver the maximum statistical power, arguably the sample does not represent the underlying population by undersampling active firms (in practice, there are far more active firms than failed firms). If we estimate a model, say, logit model based on such a matched sample, anything special we need to consider when interpreting the results? Thanks.
    Tim

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats